K. Zhu et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 1. Synthesis of 4a–4c. Reagents and condi-
tions: a) triethylamine, dichloromethane, 0 °C to r.t.,
overnight, 62–84%; b) 1,2,3,4-tetrahydroisoquino-
line, triethylamine, acetonitrile, microwave heating
100 °C, 10 min, 66%-83%.
Scheme 2. Synthesis of 4b1–4b20. Reagents and
conditions: a) triethylamine, dichloromethane, 0 °C
to r.t., overnight, 57%–85%; b) 1,2,3,4-tetra-
hydroisoquinoline, triethylamine, acetonitrile, mi-
crowave heating 100 °C, 10 min, 42%–79%.
Design and synthesis PRMT5 inhibitors
withdrawing group at 3-position had decreased activity or no obvious
inhibition on PRMT5. Since compounds 4b4 and 4b9 exhibited in-
creased activity, analogue 4b14 with 3-Br and 2-CN was next prepared.
As expected, 4b14 showed the most potent activity with an IC50 value
Compound 17 and GSK-3326595 were previously reported PRMT5
inhibitors. While 17 was identified and proposed to occupy the SAM
binding site in our recent work,16 GSK-3326595 was confirmed as a
substrate binding site inhibitor.9 A detailed analysis of their binding
modes allowed the identification of key pharmacophores in each
compound, and the following anticipation of the merging of these
fragments including tetrahydroiso-quinoline (blue color) and methyl 2-
acetamidobenzoate (red color). In order to investigate the optimal
distance between both fragments for desired activity, three alkyl linker
(black color) was first rationally exploited to yield the new potential
PRMT5 inhibitors 4a–4c (Fig. 2).
of 2.71
0.21 µM, being 11.7-fold as active as the initial hit 4b.
Further introduction of substituents on the 4-position of benzene
(4b15–4b17) did not have positive effects on the PRMT5 inhibitory
activity. In addition, replacement of the benzene ring with bulky
naphthalene or quinolone moieties also led to the loss of activity.
Selectivity test for 4b14
through acylation reaction of methyl 2-aminobenzoate with corre-
sponding acyl chlorides 2a–2c. Then 3a–3c were reacted with 1,2,3,4-
tetrahydroisoquinoline under microwave condition to give corre-
sponding target compounds 4a–4c in 66%–85% yields. The three
compounds were then evaluated for PRMT5 inhibitory activity at en-
zymatic level. To investigate the substituent effect of phenylamine
fragment, twenty analogues of 4b were synthesized. Scheme 2 showed
materials 2b and 5b1-5b20 in 33%-58% overall yields, which was
achieved according to the same protocol as that of 4a–4c.
In enzyme-inhibitory assays using PRMT1 and PRMT4, compound
4b14 was assessed for its activity against the two enzymes to confirm its
selectivity. As shown in Fig. S1 (Supporting information), the IC50 va-
lues of 4b14 against PRMT1 and PRMT4 were both above 200 μM,
which indicated that 4b14 was a selective inhibitor of PRMT5 over
PRMT1 and PRMT4.
Mechanism of action and binding mode analysis
The direct binding of compound 4b14 to PRMT5 was confirmed by
surface plasmon resonance (SPR) assay with a KD (equilibrium dis-
sociation constant) value of 10.8
3.2 μM (Fig. 3A). The mechanism
PRMT5 inhibitory activity evaluation and structure-activity
relationship analysis
of action (MOA) of 4b14 was assessed by determining the IC50 values in
the presence of various concentrations of SAM and peptide. As shown in
Fig. 3B, 4b14 displayed competitive inhibition with peptide while
substrate binding site inhibitor of PRMT5. To probe the probable
binding mode of 4b14 with PRMT5, molecular docking and molecular
dynamics simulation were performed. The putative binding mode and
the detailed interactions between 4b14 and PRMT5 were shown in
Fig. 3C and D, respectively. As indicated in Fig. 3D, although there are a
cation-π interactions as expected.
Radiometric-based scintillation proximity assay (SPA) was used to
detect the inhibitory activity of the synthesized 4a–4c against PRMT5.
The bioassay results showed that compound 4b displayed better PRMT5
inhibitory activity (IC50 = 31.62
2.15 µM) than 4a and 4c, in-
dicating that propionyl group (e(CH2)2COe) is the superior linker for
the improved activity.
To expand the substituent/fragment diversity and investigate elec-
tronic or steric effect on improving the PRMT5 inhibitory activity,
compound 4b was next used as a template based on the above bioassay
result. The introduction of different substituents into phenylamine
moiety or replacement of phenylamine moiety by other fragments were
carried out with chemical agents in hand, which resulted into the
production of 4b1–4b20. Among the compounds (4b1–4b4) with ortho-
substituents, 4b4 with 2-CN showed slightly better activity
Based on this binding mode, 100 ns molecular dynamics (MD) si-
mulation were performed to further validate the putative binding mode
and calculate the binding free energy. As shown in Fig. S2 (Supporting
information), the interactions between 4b14 and PRMT5 were found to
be stable during the MD simulation, by analyzing the root mean square
deviation (RMSD) values of PRMT5 (heavy atoms Cα, C, N) and 4b14,
respectively. Then the binding free energy was obtained by MM/PBSA
method to be −36.38 kcal/mol, which was well in accordance with its
enzymatic inhibition activity.
(IC50 = 25.55 1.67 µM) compared with 4b. From the assay results of
4b6–4b13, compound 4b9 with 3-Br showed the best activity
(IC50 = 18.12 0.71 µM), while others with more bulky or electron-
3